Helen Frowe’s Defensive Killing is in many respects an excellent book, full of arguments that are original, interesting, important, and often persuasive. In other respects, the book is deeply unsettling, as it forcefully challenges the belief that killing ordinary civilians in armed conflict is a paradigmatic moral wrong. In particular, Frowe argues that civilians who make political, material, strategic, or financial contributions to an unjust war may lose their moral protection from intentional and collateral harm. On this point, Frowe’s arguments are original, interesting, and important but, thankfully, not persuasive.
1 Coming from the Mill
The front cover of Helen Frowe’s Defensive Killing features a painting by LS Lowry entitled Coming from the Mill. The painting depicts a number of British workers as they leave a Manchester factory. They exit onto a busy street full of ordinary people going about their ordinary lives. Nearby, five children are at play.
Just below the painting, the title of the book appears, in lower case, the word ‘killing’ in red. The image and the title may seem mismatched, but in fact they go together. According to Frowe, such ordinary people going about their ordinary lives are often morally liable to defensive killing in war.
The notion that one might defensively kill ordinary civilians in war may seem strange. In international law, civilians enjoy general legal protection from intentional, unnecessary, or disproportionate harm.1 Civilians lose this general legal protection only for such time as they take a direct part in hostilities.2 Typically, direct participation in hostilities is construed narrowly, to include performing specific acts, or participating in military operations, likely to directly cause serious harm.3 Accordingly, civilians retain their general legal protection even if they contribute to the general war effort or engage in war-sustaining activities including political advocacy, voting, producing food, and paying taxes. Although Frowe accepts these legal rules, largely for pragmatic and epistemic reasons, she argues that they lack a firm basis in deep moral principles.
I will refer to civilians who make such indirect contributions to an unjust war as contributing civilians.
make political contributions, such as voting for the war, writing pro-war articles or books, attending pro-war rallies, and so on. They make material contributions by producing weapons, equipment, and food. … Non-combatants also make technological contributions to war, perhaps by designing or testing weapons and machinery for military use. They can also make what we can call strategic contributions, for example by settling land … . And, of course, the war is financed by taxes paid by combatants and non-combatants alike.4
Frowe argues that contributing civilians are indirect threats, by which she means that they make causal contributions to the direct threats posed by their armed forces.5 Frowe writes that “[t]urning oneself into … part of the causal chain that kills [an innocent person]” can make one morally liable to be killed.6 In particular, indirect threats are liable to defensive killing if they intentionally fail to take reasonable opportunities to avoid contributing to unjust direct lethal threats. Accordingly, Frowe proposes that contributing civilians are morally liable to be killed if they intentionally fail to take reasonable opportunities to avoid contributing to the unjust lethal threats posed by their armed forces.
In my view, many contributing civilians may have no reasonable opportunities to avoid contributing to the unjust lethal threats posed by their armed forces. In most countries, tax evasion is punished either by imprisonment or by heavy fine. Many employers withhold their employees’ income tax liability from their paychecks. Most goods, including food and clothes, carry a sales tax. Arguably, contributing civilians who pay income and sales taxes in order to avoid criminal punishment, keep their job, or feed and clothe themselves are justified in doing so. No contributing civilian makes anyone worse off by paying her taxes and any contributing civilian who fails to pay her taxes will be made much worse off. Though the war to which she contributes is unjust, her contribution seems justified and not merely excused. I will argue that these civilians are not morally liable to defensive killing on any reasonable interpretation of “reasonable opportunity.”
At the same time, many contributing civilians seem to have reasonable opportunities to avoid making political, material, or strategic contributions to an unjust war. For example, civilians generally suffer no penalty if they do not vote for war-making leaders, advocate war, or work in tank factories as opposed to car factories. I will argue that these contributing civilians are also not liable to defensive killing. Their moral duties to avoid contributing to an unjust war are not sufficiently stringent for their breach of those duties to trigger the loss of their right not to be killed.
Before we proceed, a few more stipulations are in order. First, ‘[a] threat of harm is unjust if it is aimed at someone who possesses a right not to suffer the harm.’7 A person is liable to defensive killing if she lacks a right not to be killed to prevent an unjust threat to another person. Typically, she lacks such a right because she has somehow forfeited her usual right not to be killed for this purpose. An unjust war is a military campaign that lacks a just cause, that is unnecessary to secure its just cause, or that is disproportionate to its just cause. A person is eliminatively harmed if she is harmed to prevent her presence from making others worse off than they would be in her absence. A person is opportunistically harmed if she is intentionally or foreseeably harmed in the course of deliberately taking advantage of her presence to make others better off than they would be in her absence.8
Finally, a note on methodology. Frowe presents her arguments through a series of colorful hypothetical cases meant to elicit the reader’s moral intuitions. Although some have doubts about this approach, I will adopt it for present purposes.
2 Contributing Civilians as Indirect Threats
As noted above, Frowe argues that a person is morally liable to defensive killing if she poses or contributes to an unjust lethal threat while intentionally failing to take reasonable opportunities to avoid doing so. Whether an opportunity to avoid posing or contributing to a threat is reasonable or unreasonable depends on ‘several considerations, including the cost to the agent of taking the opportunity, the prospective unjust harm to [the] Victim if she does not take the opportunity, the sort of threat that the agent poses, and whether the agent will be allowing herself to threaten or causing herself to threaten.’9
Significantly, Frowe argues that ‘one ought to bear lethal costs rather than cause oneself to non-defensively kill an innocent person,’ on the grounds that doing harm is morally much worse than allowing harm.10 Put another way, doing harm is harder to justify, or subject to a more stringent negative duty, than allowing harm. Frowe does not defend the distinction between doing harm and allowing harm at length, but instead cites an earlier article in which she does so.11 In that article, Frowe argues that doing harm typically makes the victim worse off than she would have been in your absence, while allowing harm leaves the victim no worse off than she would have been in your absence.
Crucially, Frowe also argues that one ought to bear lethal costs rather than ‘[t]urnoneself into an indirect cause – into part of the causal chain that kills [a] Victim.’12 It is on this basis that Frowe later concludes that if a civilian ‘has a reasonable opportunity to avoid posing an [indirect] unjust threat and intentionally fails to take that opportunity, she is liable to defensive killing.’13 According to Frowe, civilians pose indirect unjust threats by causally contributing to the direct unjust threats posed by their armed forces.14 Frowe concludes that civilians must avoid causally contributing to unjust threats, even at very high cost to themselves, or make themselves morally liable to defensive killing.
Frowe writes that “I do not think that Driver is permitted to drive the car. He ought to let himself be shot rather than help bring about the shooting of an innocent person. Even though this is a very high cost to bear, it is a reasonable cost because the alternative is, again, to help inflict an equally costly, non-defensive harm on an innocent person.” Frowe submits that Driver must not drive the car, even to avoid his own death, and is therefore morally liable to defensive killing if he does so.
Drive-By: Terrorist holds a gun to Driver’s head, and orders him to drive the car for a drive-by shooting, in which Terrorist will kill Victim.15
I agree with Frowe that Driver is morally liable to defensive killing only if he is morally required to allow himself to be killed rather than help Terrorist. Is Driver so morally required? I am unsure. However, I am quite sure that this case does little to establish that contributing civilians are liable to defensive killing.
In this variation, Driver’s participation in the planned killing seems superfluous, making no necessary contribution to the unjust threat posed by Terrorist and making no counterfactual difference to whether Victim lives or dies. Instead, Driver’s participation only makes it $20 cheaper for Terrorist to kill Victim. Driver benefits Terrorist, but not at Victim’s expense.
Uber: Same as Drive-By except that if Driver refuses, then Uber will happily drive Terrorist to kill Victim in exchange for $20.
Is Driver morally prohibited from driving Terrorist, even to save his own life, if doing so will change nothing except the monetary cost to Terrorist? Is Driver morally liable to defensive killing if he helps Terrorist? Possibly, though any intuition regarding Driver’s liability seems far weaker than common intuitions about civilian immunity.
Arguably, in Uber, if Driver drives Terrorist to kill Victim then Driver will not merely enable Terrorist to kill Victim but will in fact jointly kill Victim with Driver. In order to live, Driver must approach Victim quickly enough to get close to Victim but not so quickly to frighten Victim away, must slow down to enable Terrorist to aim and fire, and so forth. It may appear that Driver and Terrorist are exercising joint agency, acting together to achieve a shared aim, integrating their roles in a common plan, meshing sub-plans as needed.16 If Terrorist and Driver together kill Victim through such a coordinated joint action, then Driver kills Victim even if Terrorist pulls the trigger. Arguably, Driver may not jointly kill Victim to avoid being killed by Terrorist. Importantly, Uber does not show that contributing to an unjust threat generates liability to defensive killing, although it may show that jointly posing an unjust threat generates liability to defensive killing.
Of course, contributing civilians do not jointly pose threats together with their armed forces, any more than they jointly clean streets together with their sanitation department, jointly educate children together with public school teachers, or jointly put out fires together with firefighters. There are borderline cases of joint agency, but these are not among them. Civilians do not jointly pose unjust threats unless and for such time as they directly participate in hostilities. Accordingly, Uber does not support the liability of contributing civilians to defensive killing.
In this variation, Driver would not jointly kill Victim together with Terrorist, but would make a necessary contribution to Terrorist individually killing Victim.
Carjack: Same as Drive-By, except that Terrorist will kill Driver unless Driver gives Terrorist the car, which Terrorist needs to kill Victim by himself.
Is Driver morally required to allow oneself to be killed rather than give up his car? Is Driver liable to defensive killing if he does so? I find that hard to believe. True, by giving Terrorist his car, Driver would make Victim worse off. However, Driver does not threaten to kill Victim directly, jointly with Terrorist, or indirectly though an agent that Driver effectively controls. Even if Victim could save herself by killing Driver, it seems artificial to say that she would thereby defend herself from Driver. Driver poses no threat to Victim, though he may enable Terrorist to pose a threat to Victim.
Let us stipulate that Fight Fan does not act jointly with Bully (for example, because Bully does not know who tossed him the broken bottle or why they did so) or act through Bully by using Bully as his agent (for example, because Fight Fan can affect Bully’s capabilities but cannot shape Bully’s intentions). Thus, in my view, Fight Fan does not pose an unjust threat to Nice Guy but merely enables Bully to do so.
Bar fight: Bully attacks Nice Guy with a broken bottle for no good reason. When Nice Guy disarms Bully, Fight Fan tosses Bully another broken bottle, hoping to see the two strangers fight to the death.
Suppose that, in Bar Fight, Nice Guy could kill Fight Fan before Fight Fan tosses the broken bottle to Bully. This killing would be eliminative, since it would prevent Fight Fan’s presence from making Nice Guy worse off than he would be in Fight Fan’s absence. In addition, Fight Fan could have easily avoided making Nice Guy worse off in the first place, simply by not enabling an unjust threat. Fight Fan could also have easily avoided the need for either Fight Fan or Nice Guy to be harmed, simply by not enabling an unjust threat.17 These facts also seem to reduce the wrongfulness of killing Fight Fan, perhaps to the point of making it permissible to do so all things considered.
In this variation, Driver would neither jointly pose a threat to Victim together with Terrorist nor make a necessary contribution to the threat individually posed by Terrorist. Is Driver morally required to allow himself to be killed rather than give Terrorist the car? Is Driver liable to defensive killing if he does so? I find that very hard to believe.
Uber/Carjack: Same as Drive-By, except that if Driver does not give Terrorist the car then Uber will drive Terrorist to kill Victim for $20.
To conclude, Frowe has not shown that we are morally required not to superfluously contribute to direct threats posed by others at great cost to ourselves. In these cases, we pose no unjust threat directly, jointly with others, or indirectly through agents we control. Moreover, our contribution will not make anyone worse off and our non-contribution will make us much worse off.
Almost all contributing civilians make superfluous contributions to the direct threats posed by their armed forces. There may be exceptions, such as nuclear scientists or expert bomb-makers, who make a unique and indispensible contribution to the threats posed by their armed forces. However, Drive-By fails to show that a significant number of contributing civilians are liable to defensive killing.
3 Intervening Agency and Causal Thresholds
Having made her prima facie case that contributing civilians are liable to defensive killing, Frowe turns to consider alternative views.18
Frowe rejects Rodin’s principle, citing the following case:
holds that if two or more agents create an unjust threat between them, and both meet a threshold of minimal moral responsibility, only the agent whose action came last in the causal chain counts as engaged in an unjust attack. Rodin further argues that only this agent – the proximate cause – is a permissible target of defence.19
In this case, Mayor would be the direct cause of Citizen’s death, yet Nazi seems morally liable to defensive killing. Mayor’s intervening agency does not seem to relieve Nazi of his moral responsibility for contributing to the unjust threat. Frowe concludes that ‘intervening agency … doesn’t give us a plausible way to sustain non-combatant immunity.’21
Coercion: The German army capture[s] a town. Nazi forces Mayor to shoot Citizen, on pain of Nazi shooting a much greater number of citizens.20
In this case, Cheerleader’s encouragement makes a superfluous contribution to the threat posed by Killer, one that does not make Victim worse off. It is hard to believe that Cheerleader is liable to defensive killing, even if this would somehow save Victim from Killer. Since almost all contributing civilians make individually superfluous contributions to the direct threats posed by their armed forces, Coercion does not support the liability of contributing civilians to defensive killing.
Cheerleader: Killer is about to kill Victim. Cheerleader yells “Go get him! I hate that guy!” Cheerleader intends for his encouragement to bolster Killer’s resolve, and give Killer an additional reason to kill Victim. Killer takes some comfort from Cheerleader’s encouragement, but would attempt to kill Victim regardless.
My own view is that Nazi will not merely contribute to Mayor killing Citizen or merely enable Mayor to kill Citizen. Instead, Nazi will kill Citizen through or by means of Mayor. In criminal law terms, Nazi is not a mere accomplice but an indirect perpetrator.22 It is not as if Mayor conditionally intends to kill Citizen and Nazi merely helps Mayor by fulfilling the relevant condition or by providing a necessary means or opportunity. Instead, it is up to Nazi to tell Mayor what to intend. Of course, Mayor is a moral agent who could choose to defy Nazi and spare Victim. However, so long as Mayor remains committed to saving the much greater number of citizens, Nazi can manipulate Mayor’s commitment as if it is a causal mechanism, rationally requiring Mayor to kill on Nazi’s command.23 It follows that, on my view, Coercion does not show that causal proximity and intervening agency are irrelevant to liability. Instead, Coercion shows that, if one person effectively controls another, then the agency of the person controlled does not ‘intervene’ in the morally relevant sense. Since individual civilians do not exercise control over the conduct of their armed forces, Coercion does little to undermine civilian immunity.
Frowe submits that ‘Victim would be permitted to kill any individual member of the mob (or even all of them)’ to defend himself.26
Hit: Mafia Boss wants to take Victim out, but he cannot afford to hire Assassin, who is extremely skilled and thus extremely expensive. Mafia Boss has a whip-round amongst all the members of his mob, none of whom really like Victim. Everyone coughs up a few pounds for the assassination fund.25
In my view, Hit suggests what international criminal lawyers call indirect co-perpetration, which combines joint and indirect perpetration.27 It seems that the members of the mob are working together to kill Victim. Mafia Boss’s role in the common plan is to hire Assassin and order Assassin to kill Victim. On this interpretation, the members jointly threaten to kill Victim indirectly, through an agent they control, and are therefore morally liable to defensive killing.
In this variation, each business owner superfluously contributes to Mafia Boss killing Victim through Assassin. No business owner makes a necessary contribution, since Mafia Boss could compensate for the non-payment of any owner by asking more from others or by paying more himself.
Another Hit: Mafia Boss wants Victim dead, but he cannot afford to hire Assassin, who is extremely skilled and thus extremely expensive. Mafia Boss demands money from local business owners, on pain of death, serious injury, or destruction of their business. Everyone pays a few pounds for the assassination fund.
In my view, it is unreasonable to demand that any business owner refuse to superfluously contribute. No individual contribution will make Victim worse off and any owner’s refusal to contribute would make her much worse off.28 Since contributing civilians almost always make individually superfluous contributions to threats directly posed by their armed forces (and indirectly posed by their governments), Another Hit does not support the liability of contributing civilians to defensive killing.
One might object that, unlike in Hit, the business owners in Another Hit do not intend for Assassin to kill Victim. Of course, we could alter the case so that Mafia Boss threatens to harm the business owners if Assassin fails to kill Victim. In that variation, the business owners may very well intend for Assassin to kill Victim. In any event, Frowe’s own view is that “[k]nowledge, even absent intention, can be sufficient to ground liability.”29
It is true that, in Another Hit, the business owners collectively enable Mafia Boss to hire Assassin and thereby threaten Victim. Although each individual contribution is superfluous, the sum of their individual contributions is necessary. In my view, this fact makes no difference to their individual liability to defensive killing. As I see it, one person’s liability to defensive killing should depend on what they do, not on what others do. Although liability can depend on what one person jointly does, together with others, it cannot depend on what others do independently.
More precisely, only a person’s own actions can trigger or ground their liability. The independent actions of others can only affect constraints or limiting conditions on their liability. For example, the independent actions of others may affect the necessity of using defensive force, and necessity may be an internal constraint on liability. But the ground of liability remains what an individual does directly, jointly with others, or indirectly through others she effectively controls.
In Another Hit, the business owners do not jointly enable Mafia Boss to hire Assassin to kill Victim. Business owner 1 independently pays, business owner 2 independently pays, and so on. Whether business owner 1’s payment contributes to an unjust threat depends on what business owners 2 through N independently do. Yet, in my view, business owner 1’s liability to defensive killing should not depend on what business owners 2 through N independently do.
The business owners in Yet Another Hit jointly enable an unjust threat to Victim. Are they morally liable to defensive killing? I think not. In response to Carjack, I suggested that individually enabling an unjust threat does not make one liable to defensive killing. Unsurprisingly, I do not believe that jointly enabling others to pose unjust threats, without more, makes one liable to defensive killing. The permissibility of killing those who enable unjust threats depends on other considerations – including whether they could have easily avoided doing so and thereby easily avoided being in a situation in which someone will be harmed – as well as the number of such persons who would be killed as compared to the number of innocent persons who would be saved.
Yet Another Hit: The same as Another Hit, except that Mafia Boss assembles the business owners, tells them how much money he needs, and tells them to divide up the costs among themselves. The business owners deliberate and come up with a payment plan. Each business owner performs his or her role in the plan.
In any event, contributing civilians do not jointly enable their armed forces to pose unjust threats. For example, whether one person’s vote contributes to an unjust war depends on how millions of others independently vote. If I drive you to the polling station then perhaps we jointly vote. But we do not jointly vote with all or even most of our fellow citizens. I vote. You vote. They vote. It is hard to see why your liability to be killed should depend on how millions of other people independently vote. More generally, the uncoordinated, individually superfluous contributions of many people – whether financial, political, or material – seem inapt to make any of them morally liable to defensive killing.
To be clear, we can have duties to make or refrain from making individually superfluous contributions to certain outcomes, including to political and environmental outcomes. These duties can become fairly stringent when grounded in considerations of fair play, gratitude, solidarity, or expressive value. However, in my view, such duties are not so stringent that their violation will result in moral liability to lethal harm. It is to the relationship between stringency and liability that we now turn.
4 Stringency and Motivation
As we have seen, Frowe argues that a person may be morally liable to defensive killing if she had an opportunity to avoid contributing to an unjust lethal threat at a morally required, non-lethal cost but did not take that opportunity. Put another way, the unjustified breach of a moral duty can ground liability to defensive killing even if one could justifiably breach that same duty in order to avoid one’s own death.
For example, suppose that the business owners in Another Hit are not threatened with retaliation but instead volunteer to contribute to the assassination fund simply because they dislike Victim. On Frowe’s view, the owners are liable to defensive killing even though they are not morally required to allow themselves to be killed rather than contribute.
According to Frowe,
Selfish Bridge: Victim is fleeing Murderer, who wants to kill him. Victim’s only escape route is across a narrow bridge that can hold only one person. Selfish Pedestrian is out for a walk on the bridge. Selfish Pedestrian could easily get off the bridge, but doing so will involve getting her feet wet and she has on her lovely new shoes. She decides, with a certain amount of regret, to stay on the bridge, realizing that her doing so impedes Victim’s escape.30
Frowe argues that, ‘by getting off the bridge at the cost of wet shoes, Selfish Pedestrian could have ensured that neither she nor Victim needed to bear a serious harm. In refusing to bear this reasonable cost, she becomes morally responsible for endangering Victim’s life.’
even though Selfish Pedestrian is not initially required to bear a lethal cost to avoid posing a threat to Victim … Selfish Pedestrian’s initial refusal to bear a reasonable cost  makes her liable to bear a much higher cost as part of Victim’s defence of his life.
Of course, unlike Frowe’s Selfish Pedestrian, individual contributing civilians do not pose threats to others – even in Frowe’s rather unconventional sense – since their presence makes no one worse off than would their absence. Relatedly, individual contributing civilians generally cannot ensure that neither they nor others will suffer serious harm. Except in rare cases, there is nothing that individual contributing civilians can do to avoid being seriously harmed or to prevent others from being seriously harmed. So cases like Selfish Bridge do not show that individual contributing civilians are liable to defensive killing.
In my view, it would be wrong for Peter to kill Burns to save Tim. Although Burns culpably breaches his moral duty to save Tim, that moral duty is not sufficiently stringent for its breach to render Burns liable to defensive killing. Certainly, Burns is not morally required to cause his own death to save Tim. Nor does Burns seem morally required to allow himself to be killed to save Tim. Accordingly, it would be unjust to opportunistically harm Burns to save Tim, inflicting more cost on Burns than he is required to accept to save Tim.33
Pond: Tim is drowning in a pond. Burns could easily save Tim at little cost to himself, but chooses not to do so. Peter cannot save Tim himself, but can shoot Burns so that his dead body falls into the pond. Tim could then use Burns’ body as a floatation device until help arrives.32
Nevertheless, it might be all-things-considered permissible for Victim to kill Selfish Pedestrian. For one thing, killing Selfish Pedestrian would be eliminative rather than opportunistic, preventing her presence from making Victim worse off rather than taking advantage of her presence to make Victim better off. For another, Selfish Pedestrian could easily avoid making Victim worse off due to her presence, by jumping off the bridge at little cost to herself. Finally, Selfish Pedestrian could easily avoid being harmed, and the ease with which someone could avoid harm may reduce the wrongfulness of harming her. These considerations might combine to make it permissible for Victim to kill Selfish Pedestrian in order to save herself. None of these considerations apply to the killing of contributing civilians.
5 Eliminative, Opportunistic, and Exploitative Killing
Frowe submits that ‘[i]t does not seem impermissible for Victim to kill the knife-thrower if this is the only way in which he can save himself.’34 In particular, Frowe writes that ‘[s]uch merely opportunistic harming lacks the moral repugnance that attaches to the exploitative harming of bystanders to make oneself better off than one would have been in their absence.’35
Mob: Victim is being chased by an angry racist mob who[se members] are trying to kill him. They corner him in a dead-end street. One of the mob [members] throws a knife to the ringleader, who advances menacingly towards Victim. Unbeknown to the mob, Victim has a concealed gun. But he also has terrible eyesight: his short-distance vision is just awful. He cannot hope to accurately aim at the ringleader at such close range. His only hope of saving his life is to shoot the guy who threw the ringleader the knife, who is standing further away. When the rest of the mob [members] realize that he has a gun and he’s not afraid to use it, they will be scared off.
I find Mob, like Drive-By, somewhat ambiguous. If the ringleader needs the knife to kill Victim then the knife-thrower makes Victim worse off than Victim would be in the knife-thrower’s absence. On these facts, the knife-thrower can be eliminatively killed before he throws the knife or merely-opportunistically killed after he throws the knife. Either way, killing the knife-thrower prevents him from making Victim worse off than Victim would be in his absence. Frowe’s analysis seems to go through on those facts.
In contrast, if the ringleader does not need the knife to kill Victim then the knife-thrower does not make Victim worse off than Victim would be in the knife-thrower’s absence. On the contrary, killing the knife-thrower as a means of scaring off the rest of the mob will make Victim better off than Victim would be in the knife-thrower’s absence. On those facts, killing the knife-thrower would be ‘exploitative’ as Frowe defines that term.
As we have seen, the individual contributions of contributing civilians generally do not make anyone worse off. It follows that contributing civilians cannot be eliminatively killed or merely-opportunistically killed but only exploitatively killed (or gratuitously or collaterally killed, of course). Since Frowe assumes that killing the knife-thrower in Mob is not exploitative she does not even try to argue that the knife-thrower (and, by analogy, contributing civilians) might be liable to exploitative killing. It follows that Frowe’s analysis of Mob does not support the liability of contributing civilians.
On my view, the knife-thrower might be liable to defensive killing even if he does not make Victim worse off (suppose the ringleader already has a knife) on the grounds that he jointly threatens to violate Victim’s rights together with the other mob members. Since contributing civilians do not jointly violate anyone’s rights together with their armed forces, my analysis of Mob does not support the liability of contributing civilians to exploitative killing. As I have shown, neither does Frowe’s.
6 Facilitating and Triggering
Frowe submits that Victim has a moral duty to allow Bully to pinch her rather than trigger Bully to kill five people because Victim would have a moral duty to suffer a comparable harm to save five people from being killed. In contrast, if Bully tries to kill Victim then Victim is morally permitted to defend herself even if doing so will so enrage Bully that he will kill five people. This is because Victim has no moral duty to allow herself to be killed in order to save five people from being killed and, on Frowe’s view, triggering harm is morally comparable to allowing harm. Notice that, in both cases, Bully will kill five people only if Victim resists, such that Victim would make the difference between the five living and the five dying. Victim’s presence threatens to make the five worse off than they would be in her absence.
Pinch: Bully wants to pinch Victim’s arm, which will hurt. Victim knows that if (and only if) he tries to prevent this, Bully will become so enraged that he will kill five people.
Returning to Drive-By, I take it that Driver has no moral duty to allow himself to be killed in order to save Victim from a lethal threat to which Driver does not contribute. It follows that, on Frowe’s view, Driver has no moral duty to allow himself to be killed rather than trigger mediated harm to Victim. Yet Frowe writes that Driver has a moral duty to allow himself to be killed rather than facilitate mediated harm to Victim. On this view, triggering mediated harm is morally comparable to failing to save while facilitating harm is morally comparable to doing harm.
What justifies such a sharp moral asymmetry between triggering and facilitating foreseen but unintended mediated harms? Again, we are assuming that the harms will occur only if they are triggered or facilitated, as the case may be. Why does one have a moral duty to allow oneself to be killed rather than make a necessary contribution to a mediated harm but no moral duty to allow oneself to be killed rather than provide the necessary trigger for a mediated harm?
In a footnote, Frowe writes that ‘this way of ‘causing’ harms [that is, triggering harms] is very different from the facilitating of harms… as [when] Driver helps Terrorist kill Victim in Drive-By.’37 Perhaps it is, but Frowe gives us no reason to think so. It follows that Frowe has given us no reason to think that enabling others to pose unjust threats renders one liable to defensive killing.
Lowry painted Coming from the Mill, in 1930, at a time when Britain was not at war, although it enforced colonial rule through the threat of military force. Perhaps the ordinary civilians that the painting depicts would not be morally liable to defensive killing on Frowe’s account.
However, had Lowry painted a similar scene a few decades earlier, during the Irish War of Independence (1919–1921), the Third Anglo-Afghan War (1919), or the Third Anglo Marri war (1917), then perhaps its subjects would be morally liable to defensive killing on Frowe’s account. Yet the scene that Lowry depicts, though hardly timeless, is not tied to any particular year or decade. It is a familiar scene of urban life in the industrial age, teeming with ordinary people striving to hold down a job, to raise a family, and to savor whatever happiness they manage to secure. It is hard to believe that their most basic right, their right not to be killed, depends on what war their government happens to be fighting when they happen to come of age.
Helen Frowe’s Defensive Killing is in many respects an excellent book, full of arguments that are original, interesting, important, and often persuasive. In other respects, the book is deeply unsettling, as it forcefully challenges the belief that killing ordinary civilians in armed conflict is a paradigmatic moral wrong. On this point, Frowe’s arguments are original, interesting, and important but, thankfully, not persuasive.
Adil Ahmad Haque is a member of the Associate Graduate Faculty of the Rutgers University Department of Philosophy and co-Director of the Rutgers Institute for Law and Philosophy. He serves on the editorial boards of Law and Philosophy, Criminal Law and Philosophy, and Criminal Law and Criminal Justice Books. His first book, Law and Morality at War, was recently published by Oxford University Press.
Thanks to Seth Lazar for comments on an earlier draft.
Protocol Additional to the Geneva Conventions of August 12, 1949, and relating to the Protection of Victims of International Armed Conflicts, June 8, 1977 (Protocol i) art 51(1). See also International Committee of the Red Cross, Customary International Humanitarian Law, vol. 1 (cup 2009) p. 19.
Protocol i art 51(3); icrc, Customary ihl, p. 19.
International Committee of the Red Cross, Interpretive Guidance on the Notion of Direct Participation in Hostilities Under International Humanitarian Law (icrc 2009) p. 46.
Helen Frowe, Defensive Killing (oup 2014) p. 164.
Ibid pp. 7, 15–16.
Ibid p. 80.
Frowe, p. 72 (emphasis added).
See, e.g., Warren S Quinn, ‘Actions, Intentions, and Consequences: The Doctrine of Double Effect’ (1989) 18 Philosophy & Public Affairs 334; Victor Tadros, ‘Wrongful Intentions without Closeness’ (2015) 43 Philosophy & Public Affairs 52.
Frowe, Defensive Killing, p. 11.
Ibid p. 79.
Helen Frowe, ‘Killing John to Save Mary: A Defence of the Moral Significance of the Distinction between Doing and Allowing,’ in Joseph Keim Campbell, Michael O’Rourke, and Harry S. Silverstein (eds.), Action, Ethics, and Responsibility (Cambridge, Mass.: mit Press, 2010), 47–66.
Frowe, Defensive Killing, p. 80. Ibid p. 11 (‘The threshold at which an opportunity becomes unreasonable is set highest with respect to people who will be causing themselves to directly threaten or to indirectly cause a direct threat. People who will pose these types of threat ought to bear even lethal harm rather than kill Victim’); ibid p. 172 (‘people are required to bear very high costs rather turn themselves into threats, especially when they will be indirect causes of harm’).
Frowe, Defensive Killing, p. 162. Recall that, according to Frowe, ‘A direct threat is someone who will inflict harm upon you. An indirect threat is someone who endangers you in some other way, for example by blocking an escape route that you need to take, or making some other causal contribution to what threatens you.’ In particular, ‘being a cause of a direct threat is particularly morally significant.’ Ibid.
For example, according to Frowe, ‘non-combatants knowingly turn themselves into threats—and into causes—when they produce equipment for their country’s war effort.’ Ibid p. 172.
Frowe, Defensive Killing, p. 80. Frowe later offers a variation of this case in which Driver mistakenly believes that Victim poses an unjust threat and that the drive-by shooting is a permissible exercise of defensive force. Ibid p. 84.
See Michael E Bratman, Shared Agency: A Planning Theory of Acting Together (oup 2014).
On the moral significance of the ability to avoid harm or the occasion for harm, see generally T.M. Scanlon, What We Owe to Each Other ch 6 (Harvard 1998); Victor Tadros, The Ends of Harm ch 8 (oup 2011).
Frowe, p. 162.
Frowe, p. 166.
Ibid p. 168. See David Rodin, War and Self-Defense (oup 2002) p. 63.
Frowe, Defensive Killing, 169.
See Stefano Manacorda & Chantal Meloni, ‘Indirect Perpetration versus Joint Criminal Enterprise’ (2011) 9 Journal of International Criminal Justice 159.
Rodin’s own explanation turns on Nazi’s intention, rather than on control.
Frowe, p. 175.
Frowe, Defensive Killing, p. 176. See Cécile Fabre, ‘Guns, Food and Liability to Attack in War,’ Ethics, Vol. 120 (October 2009) 36, 61. Fabre has since modified her view. See Cécile Fabre, Cosmopolitan War (oup 2012).
Frowe, Defensive Killing, p. 176.
See Manacorda & Meloni, ‘Indirect Perpetration versus Joint Criminal Enterprise.’
What if a business owner superfluously contributes to an assassination fund because she simply dislikes Victim? As I understand Frowe’s position, such a business owner has no reasonable alternative but to contribute, even if she is not motivated by the threat of retaliation.
Frowe, Defensive Killing, 171.
Frowe, Defensive Killing, p. 76.
See Adil Ahmad Haque, Law and Morality at War ch 3 and Appendix (oup 2017).
A reviewer suggests that killing someone to use their dead body may be worse than using their body in a way that kills them. We could alter the case so that Peter can hit Burns in the head with a rock, knocking Burns unconscious. Tim could then use Burns’ body as a floatation device until help arrives, but Burns will drown to death while unconscious.
On the relationship between duties to accept harm and liability to harm, see generally Tadros, The Ends of Harm. Note that it may be permissible to inflict unintentional harm on others, that they are not required to accept, as a side-effect of preventing substantially greater harm to others.
Frowe, Defensive Killing, pp. 178–79.
Ibid 179 (italics added).
Ibid 160 (italics added).